Indoor locationing refers to a process of predicting an object's location in an indoor space. For example, a user's indoor location can be predicted by comparing data from wireless access points received at the user's mobile device against a database where mappings between the wireless access points and indoor locations are stored.
An indoor location provider may create these types of databases by conducting on-site surveys at selected locations. The surveys are used to create a model of the indoor space by scanning for locations of the wireless access points and their corresponding signal strengths at those locations. However, overtime some wireless access points may disappear, move, or become otherwise unavailable, leading to inaccurate location predictions and increased location prediction errors.